Signal to noise ratio is the key to success of remote acoustic monitoring, says Pure Technologies (Europe) managing director David Youdan. With a high enough ratio, the distinctive sound of a corroded prestressing strand fracturing, for example, can be distinguished, identified and located in real time and the owner of the structure promptly informed before any significant problem can develop.
The success of the company itself, however, has been directly linked to the growth of the internet, as Youdan confirms.
'It's really caused a significant change in the attitude of bridge owners over the last two years, simply because using the internet slashes the cost of data acquisition and transmission.
We can now monitor structures anywhere in the world from our headquarters in Canada, economically, ' says Youdan.
Originally developed about six years ago to keep watch on post-tensioned floors in high rise buildings, Pure Technology's Soundprint system has recently been validated for use on grouted post-tensioned highway structures by trials carried out by the TRL (NCE 27 May 1999).
But the same basic concept has been developed and extended into new areas, such as the monitoring of very large post-tensioned concrete pipes and suspension bridges.
'Physically inspecting the main cables of large suspension bridges for corrosion is a very complicated and expensive exercise, ' Youdan points out.
'Acoustic monitoring is a very attractive alternative in principle - we just had to show it would work.'
Initial trials took place on the 65 year old Bronx Whitestone Bridge in New York. Sensors were located on each hanger cable, and an 'enormous' signal was picked up when one of the wires in the main cable was deliberately cut. 'In practice, we showed we could locate the wire break to within 0.5m longitudinally, ' Youdan says.
'Then the client wanted to know if we could locate the break accurately within the cable itself.'
Sensors fixed around the circumference did achieve an acceptable degree of accuracy - but at a price.
'It just isn't cost effective to do anything but treat the main cables as two-dimensional, ' argues Youdan. The bridge's owners agreed, and last Christmas, Pure Technologies was awarded a long term monitoring contract. Such was the power of the signal that sensors on every second hanger were judged to be adequate. Nor was it necessary to scrape off the existing paint coating on the cables to achieve metal to metal contact, Youdan adds.
It was hangers rather than main cables that concerned the managers of the George Washington Bridge only a few kilometres away. 'And it wasn't just corrosion they were worried about, ' says Pure Technologies general manager Jack Elliott.
'They wanted to be informed about all forms of external damage, from vandalism to ship impact.'
Pure Technologies was soon able to demonstrate the capability to identify significant events, despite the relatively small cross sectional area of the bridge's hangers. With the trials just complete, the company is developing plans for a permanent system.
This will do more than send incidence reports over the internet to Calgary.
'If anything serious occurs the system will immediately trigger alarms in two other locations, by ringing mobile phones or sending e-mails, ' Elliott explains.
'We don't really expect to pick up the noise of a vandal attacking a hanger - but we've shown we could tell if he was using a hacksaw or an angle grinder.'
On a typical grouted posttensioned concrete structure, such noises - and all other irrelevant sound - would be deliberately filtered out and only the distinctive signature of a wire break would be notified to the client. As the Soundprint technology moves into new sectors, there could be other significant acoustic signatures that would be of interest to infrastructure owners.
'We're pushing back the boundaries on a site by site basis, ' Youdan says.
New approach to old technology A VERY different approach has been adopted by Physical Acoustics, part of a New Jerseybased international group. 'We listen for much higher frequency sounds, in the 150kHz range, at amplitudes several orders of magnitude lower, ' explains managing director Phil Cole.
'The basic technology was developed more than 20 years ago to monitor large glass reinforced plastic tanks designed for storing corrosive chemicals.'
These had a nasty habit of failing catastrophically without warning. Safety and pollution concerns generated an emergency research programme funded by several chemicals companies. In the end, it was found that by listening for the distinctive sounds of the 'micro-earthquakes' that occur as stresses in the GRP composite approach dangerous levels, action could be taken in time to avoid catastrophe.
'Luckily, composites give plenty of warning of potential failure, ' Cole says. 'A lot of noise is generated even at 30% to 50% of failure load.
'Nowadays a standard acceptance test for new GRP tanks is to monitor the acoustic emissions as they are filled for the first time. If there's no noise, the tank will have a long service life.'
Over the subsequent decades Physical Acoustics extended the technology into steel storage tanks, pressure vessels and pipe networks. The aerospace industry, where composites play an increasing role, also adopted acoustic emission monitoring.
Central to its success was AE's ability to detect the very first stages of a problem, long before any crack or delamination could be picked up by conventional methods.
'An extra benefit is that no paint removal is needed, ' Cole points out. 'The acoustic sensors are simply mounted on the external surfaces, between 3m and 20m apart. And there's no need to wait for quiet periods - we want to monitor the structure during normal use.'
Over most of the past decade Physical Acoustics has been working with the American Federal Highways Authority to adapt AE techniques to the monitoring of steel bridges. And for much of the same period the company has been in partnership with Cardiff University's School of Engineering, carrying out research to categorise the AE from fatigue growth in a wide range of structural steel sections and fabrications. The two projects came together recently in trials carried out for the Highways Agency on two typical steel box girder viaducts.
'These built on earlier trials on viaducts in South Wales, ' explains applications engineer Jon Watson.
'Sensors at 10m centres had been shown to perform satisfactorily - the question now was not could we locate any defects, but could we identify them as well?'
During the first trial only 17 insignificant acoustic emissions were picked up during an entire rush hour - confirmation that the box girder was in good condition and operating well within its structural limits. By contrast, more than 1M emissions were recorded during the first day of a trial on a second 40m span box girder.
'Obviously, we focus in on any areas where there are multiple emissions, ' Watson says. 'In this case there were 18 significant sources, all but one located at the end of the diagonal cross-bracing inside the box.'
A local array of sensors was installed around the loudest source. Data from these was compared to the fatigue data from Cardiff University, and a preliminary identification made.
'It gives every indication of being a 250mm long fatigue crack in the bottom weld of the box girder, ' Watson reports.
Information from strain gauges on the bridge and video monitoring of traffic flows above showed the emissions peaked when traffic contained a high proportion of HGVs and was travelling at around 95km/h.
Cardiff's analysis also suggests the crack has stopped growing.
Watson adds: 'One inspection can't tell the whole story, although if you get zero emissions you can safely say the structure has a very long fatigue life.
'But if you make regular checks you will be able to pick up any defects as they develop and be able to predict when maintenance or repair will be needed.'